INDEX
Explanations
mentions of the name "Shar" or similar variations
New Auto-Interp
Negative Logits
ych
-0.19
ej
-0.16
ech
-0.15
sch
-0.15
osis
-0.14
ech
-0.14
rum
-0.14
üne
-0.14
.effects
-0.14
ella
-0.14
POSITIVE LOGITS
pton
0.18
jah
0.17
_EOF
0.17
à¥Ģर
0.16
pen
0.16
lotte
0.16
kees
0.16
qi
0.15
ffen
0.15
оиÑĤ
0.15
Activations Density 0.006%